One of the main strengths of an ontology-based system is the intensional description of the knowledge (information) about a domain, which is used in an inferencing process as a means of implying new domain knowledge. This inferencing process is very efficiently used for the explanation how an answer to a user's query was derived, increasing the trust of a user in the answers of the query. Moreover, such a derivation tree can be treated as an evidence of the relevance of an answer for the user's query. In this paper we present an approach which combines the derivation tree with the factual information in order to support the ranking of the results of an ontology-based query. Moreover, we discuss the possibilities to use such a ranking for clustering query's results. The approach has been implemented in the Ontobroker system, a main memory deductive database system.
展开▼
